Synchronization II

Synchronization II

Hakim WeatherspoonCS 3410, Spring 2012

Computer ScienceCornell University

P&H Chapter 2.11

2

AdministriviaPizza party: PA3 Games Night• Friday, April 27th, 5:00-7:00pm• Location: Upson B17

Prelim3 Review• Today, Tuesday, April 24th, 5:30-7:30pm• Location: Hollister 110

Prelim 3• Thursday, April 26th, 7:30pm• Location: Olin 155

PA4: Final project out next week• Demos: May 14-16• Will not be able to use slip days

3

Goals for TodaySynchronization• Threads and processes• Critical sections, race conditions, and mutexes• Atomic Instructions

• HW support for synchronization• Using sync primitives to build concurrency-safe data

structures• Cache coherency causes problems• Locks + barriers

• Language level synchronization

4

SynchronizationTwo processors sharing an area of memory

• P1 writes, then P2 reads• Data race if P1 and P2 don’t synchronize

– Result depends of order of accesses

Hardware support required• Atomic read/write memory operation• No other access to the location allowed between the read and

write

Could be a single instruction• E.g., atomic swap of register ↔ memory (e.g. ATS, BTS; x86)• Or an atomic pair of instructions (e.g. LL and SC; MIPS)

5

Synchronization in MIPS Load linked: LL rt, offset(rs)Store conditional: SC rt, offset(rs)

• Succeeds if location not changed since the LL– Returns 1 in rt

• Fails if location is changed– Returns 0 in rt

Example: atomic swap (to test/set lock variable)try: MOVE $t0,$s4 ;copy exchange value LL $t1,0($s1) ;load linked SC $t0,0($s1) ;store conditional BEQZ $t0,try ;branch store fails MOVE $s4,$t1 ;put load value in $s4

6

Need it to exploit multiple processing units…to provide interactive applications …to parallelize for multicore…to write servers that handle many clients

Problem: hard even for experienced programmers• Behavior can depend on subtle timing differences• Bugs may be impossible to reproduce

Needed: synchronization of threads

Programming with Threads

7

Programming with ThreadsConcurrency poses challenges for:Correctness

• Threads accessing shared memory should not interfere with each other

Liveness• Threads should not get stuck, should make forward progress

Efficiency• Program should make good use of available computing resources

(e.g., processors).Fairness

• Resources apportioned fairly between threads

8

Two threads, one counterExample: Web servers use concurrencyMultiple threads handle client requests in parallel. Some shared state, e.g. hit counts:

• each thread increments a shared counter to track number of hits

What happens when two threads execute concurrently?

…hits = hits + 1;…

…LW R0, hitsloc

ADDI R0, r0, 1 SW R0, hitsloc

…

9

Shared counters

Possible result: lost update!

Timing-dependent failure race condition• hard to reproduce Difficult to debug

addu/sw: hits = 0 + 1

lw (0)

addu/sw: hits = 0 + 1lw (0)

T1 T2

hits = 1

hits = 0time

10

Race conditionsDef: timing-dependent error involving access to shared

state • Whether it happens depends on how threads scheduled: who

wins “races” to instruction that updates state vs. instruction that accesses state

• Races are intermittent, may occur rarely– Timing dependent = small changes can hide bug

• A program is correct only if all possible schedules are safe – Number of possible schedule permutations is huge– Need to imagine an adversary who switches contexts at the worst

possible time

11

Critical sectionsTo eliminate races: use critical sections that only

one thread can be in• Contending threads must wait to enter

CSEnter();Critical section

CSExit();

T1 T2timeCSEnter();

Critical sectionCSExit();

T1 T2

12

MutexesCritical sections typically associated with mutual

exclusion locks (mutexes)Only one thread can hold a given mutex at a timeAcquire (lock) mutex on entry to critical section

• Or block if another thread already holds itRelease (unlock) mutex on exit

• Allow one waiting thread (if any) to acquire & proceed

pthread_mutex_lock(&m);hits = hits+1;

pthread_mutex_unlock(&m);

T1 T2

pthread_mutex_lock(&m);hits = hits+1;

pthread_mutex_unlock(&m);

pthread_mutex_init(&m);

13

MutexesQ: How to implement critical section in code?A: Lots of approaches….Mutual Exclusion Lock (mutex)lock(m): wait till it becomes free, then lock itunlock(m): unlock it

safe_increment() {pthread_mutex_lock(&m);hits = hits + 1;pthread_mutex_unlock(&m)

}

14

Hardware Support for Synchronization

15

Synchronization in MIPS Load linked: LL rt, offset(rs)Store conditional: SC rt, offset(rs)

• Succeeds if location not changed since the LL– Returns 1 in rt

• Fails if location is changed– Returns 0 in rt

Example: atomic swap (to test/set lock variable)try: MOVE $t0,$s4 ;copy exchange value LL $t1,0($s1) ;load linked SC $t0,0($s1) ;store conditional BEQZ $t0,try ;branch store fails MOVE $s4,$t1 ;put load value in $s4

16

Mutex from LL and SCLinked load / Store Conditional

mutex_lock(int *m) { while(test_and_test(m)){}}

int test_and_set(int *m) {old = *m;*m = 1;return old;

}

17

Mutex from LL and SCLinked load / Store Conditional

mutex_lock(int *m) { while(test_and_test(m)){}}

int test_and_set(int *m) {LI $t0, 1LL $t1, 0($a0)SC $t0, 0($a0)MOVE $v0, $t1

}

18

Mutex from LL and SCLinked load / Store Conditional

mutex_lock(int *m) { test_and_set:

LI $t0, 1LL $t1, 0($a0)BNEZ $t1, test_and_setSC $t0, 0($a0)BEQZ $t0, test_and_set

}

mutex_unlock(int *m) {*m = 0;

}

19

Mutex from LL and SCLinked load / Store Conditional

mutex_lock(int *m) { test_and_set:

LI $t0, 1LL $t1, 0($a0)BNEZ $t1, test_and_setSC $t0, 0($a0)BEQZ $t0, test_and_set

}

mutex_unlock(int *m) {SW $zero, 0($a0)

}

20

Alternative Atomic InstructionsOther atomic hardware primitives - test and set (x86) - atomic increment (x86) - bus lock prefix (x86)

21

Alternative Atomic InstructionsOther atomic hardware primitives - test and set (x86) - atomic increment (x86) - bus lock prefix (x86) - compare and exchange (x86, ARM deprecated) - linked load / store conditional

(MIPS, ARM, PowerPC, DEC Alpha, …)

22

SynchronizationSynchronization techniquesclever code • must work despite adversarial scheduler/interrupts• used by: hackers• also: noobs

disable interrupts• used by: exception handler, scheduler, device drivers, …

disable preemption• dangerous for user code, but okay for some kernel code

mutual exclusion locks (mutex)• general purpose, except for some interrupt-related cases

23

Using synchronization primitives to buildconcurrency-safe datastructures

24

Broken invariantsAccess to shared data must be synchronized• goal: enforce datastructure invariants

// invariant: // data is in A[h … t-1]char A[100];int h = 0, t = 0;

// producer: add to list tailvoid put(char c) {A[t] = c;t++;

}

// consumer: take from list headchar get() {while (h == t) { };char c = A[h];h++;return c;

}

1 2 3

head tail

25

Protecting an invariant

Rule of thumb: all updates that can affectinvariant become critical sections

// invariant: (protected by m)// data is in A[h … t-1]pthread_mutex_t *m = pthread_mutex_create();char A[100];int h = 0, t = 0;

// producer: add to list tailvoid put(char c) {pthread_mutex_lock(m);A[t] = c;t++;pthread_mutex_unlock(m);

}

// consumer: take from list headchar get() {pthread_mutex_lock(m);while(h == t) {}

char c = A[h];h++;pthread_mutex_unlock(m);return c;

}

26

Guidelines for successful mutexingInsufficient locking can cause races• Skimping on mutexes? Just say no!

Poorly designed locking can cause deadlock

• know why you are using mutexes!• acquire locks in a consistent order to avoid cycles• use lock/unlock like braces (match them lexically)

– lock(&m); …; unlock(&m)– watch out for return, goto, and function calls!– watch out for exception/error conditions!

P1: lock(m1);lock(m2);

P2: lock(m2);lock(m1);

27

Cache Coherencycauses yet more trouble

28

Remember: Cache CoherenceRecall: Cache coherence defined...Informal: Reads return most recently written valueFormal: For concurrent processes P1 and P2

• P writes X before P reads X (with no intervening writes) read returns written value

• P1 writes X before P2 reads X read returns written value

• P1 writes X and P2 writes X all processors see writes in the same order

– all see the same final value for X

29

Relaxed consistency implicationsIdeal case: sequential consistency• Globally: writes appear in interleaved order• Locally: other core’s writes show up in program order

In practice: not so much…• write-back caches sequential consistency is tricky• writes appear in semi-random order• locks alone don’t help

* MIPS has sequential consistency; Intel does not

30

Acquire/releaseMemory Barriers and Release Consistency • Less strict than sequential consistency; easier to build

One protocol:• Acquire: lock, and force subsequent accesses after• Release: unlock, and force previous accesses before

P1: ...acquire(m);A[t] = c;t++;release(m);

P2: ...acquire(m);A[t] = c;t++;unlock(m);

Moral: can’t rely on sequential consistency(so use synchronization libraries)

31

Are Locks + Barriers enough?

32

Beyond mutexesWriters must check for full buffer

& Readers must check if for empty buffer• ideal: don’t busy wait… go to sleep instead

char get() {acquire(L);char c = A[h];h++;release(L);return c;

}

headlast==head

empty

while(empty) {}

33

Beyond mutexesWriters must check for full buffer

& Readers must check if for empty buffer• ideal: don’t busy wait… go to sleep instead

char get() {acquire(L);char c = A[h];h++;release(L);return c;

}

char get() {acquire(L);while (h == t) { };char c = A[h];h++;release(L);return c;

}

char get() {while (h == t) { };acquire(L);char c = A[h];h++;release(L);return c;

}

Dilemma: Have to check while holding lock,

headlast==head

empty

34

Beyond mutexesWriters must check for full buffer

& Readers must check if for empty buffer• ideal: don’t busy wait… go to sleep instead

char get() {acquire(L);char c = A[h];h++;release(L);return c;

}

char get() {acquire(L);while (h == t) { };char c = A[h];h++;release(L);return c;

}

Dilemma: Have to check while holding lock,but cannot wait while hold lock

35

Beyond mutexesWriters must check for full buffer

& Readers must check if for empty buffer• ideal: don’t busy wait… go to sleep instead

char get() {do {

acquire(L);empty = (h == t);if (!empty) {

c = A[h];h++;

}release(L);

} while (empty);return c;

}

36

Language-level Synchronization

37

Condition variablesUse [Hoare] a condition variable to wait for a

condition to become true (without holding lock!)

wait(m, c) : • atomically release m and sleep, waiting for condition c• wake up holding m sometime after c was signaled

signal(c) : wake up one thread waiting on cbroadcast(c) : wake up all threads waiting on c

POSIX (e.g., Linux): pthread_cond_wait, pthread_cond_signal, pthread_cond_broadcast

38

Using a condition variablewait(m, c) : release m, sleep until c, wake up holding msignal(c) : wake up one thread waiting on c

char get() {lock(m);while (t == h)wait(m,

not_empty);char c = A[h];h = (h+1) % n;unlock(m);signal(not_full);return c;

}

cond_t *not_full = ...;cond_t *not_empty = ...;mutex_t *m = ...;

void put(char c) {lock(m);while ((t-h) % n == 1) wait(m, not_full);

A[t] = c;t = (t+1) % n;unlock(m);signal(not_empty);

}

39

MonitorsA Monitor is a concurrency-safe datastructure, with…• one mutex• some condition variables• some operations

All operations on monitor acquire/release mutex• one thread in the monitor at a time

Ring buffer was a monitorJava, C#, etc., have built-in support for monitors

40

Java concurrencyJava objects can be monitors• “synchronized” keyword locks/releases the mutex• Has one (!) builtin condition variable

– o.wait() = wait(o, o)– o.notify() = signal(o)– o.notifyAll() = broadcast(o)

• Java wait() can be called even when mutex is not held. Mutex not held when awoken by signal(). Useful?

41

More synchronization mechanismsLots of synchronization variations…

(can implement with mutex and condition vars.)Reader/writer locks• Any number of threads can hold a read lock• Only one thread can hold the writer lock

Semaphores• N threads can hold lock at the same time

Message-passing, sockets, queues, ring buffers, …• transfer data and synchronize

42

SummaryHardware Primitives: test-and-set, LL/SC, barrier, ...… used to build …

Synchronization primitives: mutex, semaphore, ...… used to build …

Language Constructs: monitors, signals, ...